Bickford has explained how the screw is loosened in two stages [19]. Firstly, an external force causes micromovement and slipping between the abutment screw thread and the implant, decreasing the screw preload. Secondly, if the preload value falls below a critical level, and the external force exceeds the joint separating force, then screw loosening occurs. Consequently, the higher the preload value, the greater is the resistance to screw loosening.

Many factors affect the screw preload, such as the coefficient of friction between the surface, the elastic modulus of the screw, the geometry of the screw and of the connection, the fit of the components, insertion torque, the presence of lubricant, and occlusal overload [12]. The optimal screw preload should induce a stress in the joint of about 60–75% of the yield strength of the material [20]. There are many methods to measure screw preload [21], removal torque measurement being the most common [22]. The removal torque is the rotational force used to loosen the screw. The greater the preload obtained, the higher is the removal torque value.

The literature demonstrates that humans undergo about 800,000 chewing cycles in a year [23], with the frequency of cycles ranging from 1 to 19 Hz [24]. Lundeen and Gibbs reported that the average occlusal force for fixed prostheses supported by implants is 250 N [25]. Different methods of cyclic testing have been used to simulate clinical oral function. The ISO 14801:2007 recommendations were designed for single, endosteal, transmucosal dental implants tested under “worst case” applications [26]. The implant samples protrude from a supporting resin by 3 mm, simulating the worst situation of 3 mm of vertical bone resorption around the implant. The implant samples are angled at 30° to the vertical and are stressed by vertical and oblique loads (Fig. 2).

In most studies, the removal torque values (RTVs) decreased after cyclic loading [27, 28], but in some studies, higher RTVs were reported [29, 30]. There are many independent variables that may influence the screw preload as mentioned above [12]. Paepoemsin et al. [31] evaluated the removal torque of the tapered screws and flat-head screws of an implant system and found that the RTVs reduced significantly after 1,000,000 cycles of loading. Few studies have truly focused on the effect of functional loading duration on preload maintenance of the screw. How the screw preload may change under different functional loading times is still unknown. The significant reduction of the RTVs and the patterns after cyclic loading have also not been investigated. Therefore, the purpose of the study was to investigate the RTVs’ decreasing pattern after different cyclic loading. The null hypothesis was that the RTVs are not significantly reduced after different cyclic loading.